2. Union-Find Problem Given a set {1, 2, …, n} of n elements. Initially each element is in a different set. {1}, {2}, …, {n} An intermixed sequence of union and find operations is performed. A union operation combines two sets into one.  Each of the n elements is in exactly one set at any time. FIND-SET returns the set that contains a particular element.

3. Set as a Tree S = {2, 4, 5, 9, 11, 13, 30} Some possible tree representations: 4 291130513 4 2 9 30 5 13 11 4 2 9 30 5 13

4. Result of A FIND-SET Operation FIND-SET(i) returns the set that contains element i. The requirement is that FIND-SET(i) and FIND-SET(j) return the same value iff elements i and j are in the same set. 4 291130513 FIND-SET(i) will return the element that is in the tree root.

5. Strategy For FIND-SET(i) Start at the node that represents element i and climb up the tree until the root is reached. Return the element in the root. To climb the tree, each node must have a parent pointer. 4 2 9 30 5 13 11

6. Trees With Parent Pointers 4 2 9 30 5 13 11 1 7 83226 10 20161412

7. Example 4 2 9 30 5 13 11 1 table[] 09 (Only some table entries are shown.) 4 Node Structure Use nodes that have two fields: element and parent. Use an array such that table[i] is a pointer to the node whose element is i.

8. Better Representation Use an integer array parent[] such that parent[i] is the element that is the parent of element i. 4 2 9 30 5 13 11 1 parent[] 0 1 2 3 4 5915 2913 450

9. Union Operation union(i,j)  i and j are the roots of two different trees, i  j. To unite the trees, make one tree a subtree of the other.  parent[j] = i

10. Union Example union(7,13) 4 2 9 30 5 13 11 1 7 83226 10 20161412

11. The FIND-SET Algorithm public int FIND-SET(int theElement) { while (parent[theElement] != 0) theElement = parent[theElement]; // move up return theElement; }

12. The Union Algorithm public void union(int rootA, int rootB) {parent[rootB] = rootA;} Time Complexity of union()  O(1)

13. Time Complexity of FIND-SET() Tree height may equal the number of elements in tree.  union(2,1), union(3,2), union(4,3), union(5,4) … 2 1 3 4 5 So time complexity is O(u), u = # of unions or elements in the set

14. u Unions and f FIND-SET Operations O(u + uf) = O(uf) Time to initialize parent[i] = 0 for all i is O(n). Total time is O(n + uf) We can do better!

15. Smart Union Strategies 4 2 9 30 5 13 11 1 7 83226 10 20161412 union(7,13) Which tree should become a subtree of the other?

16. Height Rule Make tree with smaller height a subtree of the other tree. Break ties arbitrarily. 4 2 9 30 5 13 11 1 7 83226 10 20161412 union(7,13)

17. Weight Rule Make tree with fewer number of elements a subtree of the other tree. Break ties arbitrarily. 4 2 9 30 5 13 11 1 union(7,13) 7 83226 10 20161412

18. Implementation Root of each tree must record either its height or the weight (i.e., the number of elements in the tree). When a union is done using the height rule, the height increases only when two trees of equal height are united. When the weight rule is used, the weight of the new tree is the sum of the weights of the trees that are united.

19. Height of a tree If we start with single element trees and perform unions using either the height or the weight rule. The height of a tree with n elements is at most  (log 2 n)  + 1. Proof is by induction. Therefore, FIND-SET(.) takes O(lgn) in the worst case.

20. Spruce up FIND-SET()  Path Compaction Make all nodes on find path point to tree root. FIND-SET(1) 121420 10 22 a 4 2 9 30 5 13 11 1 7 836 16 bc d e f g a, b, c, d, e, f, and g are subtrees Makes two passes up the tree.

21. Theorem [Tarjan and Van Leeuwen] Let T(f,u) be the time required to process any intermixed sequence of f finds and u unions. Assume that n/2 ≤ u < n a*(n + f*alpha(f+n, n) )  T(f,u)  b*(n + f*alpha(f+n, n)) where a and b are constants. These bounds apply when we start with singleton sets and use either the weight or height rule for unions and any one of the path compression methods for a find. Even though alpha() grows very slowly, we cannot consider T(f,u) as a linear function of (n+f). But, for all practical purposes, T(f,u) = O(n+f) In other words, for all practical purposes, Time complexity of FIND-SET() = O(1) Space Complexity: one node per element, O(n). Time Complexity

22. (optional slide) Time Complexity (cont.) Ackermann’s function.  A(1,j) = 2 j, j  1  A(i,j) = A(i-1,2), i  2 and j = 1  A(i,j) = A(i-1,A(i,j-1)), i, j  2 Inverse of Ackermann’s function.  alpha(p,q) = min{z  1 | A(z, p/q) > log 2 q}, p  q  1 Ackermann’s function grows very rapidly as i and j are increased.  A(2,4) = 2 65,536 The inverse function grows very slowly.  alpha(p,q) < 5 until q = 2 A(4,1)  A(4,1) = A(3,2) = A(2, A(3,1)) = A(2, A(2,2)) = A(2,16) >> A(2,4) [ A(2,2) = A(1,A(2,1)) = A(1,A(1,2)) = A(1,4) = 2 4 = 16 ] In the analysis of the union-find problem, q is the number, n, of elements; p = n + f; and u  n/2 For all practical purposes, alpha(p,q) < 5